Update README.md

README.md

@@ -19,7 +19,7 @@ license_link: https://ai.google.dev/gemma/terms
 
 # Gemma Model Card
 This model card is copied from the original [google/gemma-2b-it](https://huggingface.co/google/gemma-2b-it) with edits to the code snippets on how to run this auto-gptq quantized version of the model.
-This auto-gptq quantized version of the model had only been tested to work on cuda GPU. This quantized model utilise approximately 2.6GB of VRAM.
+This auto-gptq quantized version of the model had only been tested to work on cuda GPU and utilise approximately 2.6GB of VRAM.
 
 **Model Page**: [Gemma](https://ai.google.dev/gemma/docs)